AI Business Analyst

About the role

We are looking for an AI-focused Business Analyst who can take an opportunity from an ambiguous business problem to a tested, governed and measurable solution. You will work directly with stakeholders to understand how work is performed, identify where value is being lost, and determine whether the best response is process redesign, automation, AI or a combination.

This is not a traditional requirements-only role. You will independently lead discovery, shape and prioritise use cases, advise stakeholders, design practical solutions and build low-code agents and workflows using Microsoft technologies. You will remain involved through pilot delivery, adoption, production transition and benefits realisation.

What you'll do

Discovery and analysis

• Plan and facilitate stakeholder workshops, interviews, observation sessions and process walk-throughs.
• Map current and future processes, including decisions, hand-offs, exceptions, controls, information flows and informal workarounds.
• Translate operational needs into clear problem statements, requirements, user stories, acceptance criteria and measurable outcomes.
• Quantify baselines such as time, cost, volume, rework, response time, quality, user effort and service performance.
• Challenge weak assumptions and distinguish confirmed evidence from hypotheses requiring validation.
• Develop proportionate business cases and recommendations that compare AI with process change, automation and conventional technology.

Use-case assessment and prioritisation

• Maintain a structured opportunity pipeline from initial idea through validation, pilot, production and benefits review.
• Score opportunities against impact, feasibility, complexity, data readiness, risk, user need, repeatability and time to value.
• Estimate benefits transparently, document assumptions and avoid treating unverified productivity claims as realised value.
• Assess security, privacy, compliance, bias, transparency, operational resilience, licensing and human oversight requirements.
• Recommend which opportunities to progress, reshape, defer or stop, with clear rationale and dependencies.

Solution design and delivery

• Choose the simplest suitable solution pattern and define an achievable minimum viable product or proof of value.
• Build and configure agents using Microsoft 365 Agent Builder and Microsoft Copilot Studio.
• Create Power Automate workflows, approvals, notifications, audit trails and human hand-offs.
• Connect agents and workflows to approved Microsoft 365, SharePoint, Teams and other business data sources using supported connectors, APIs and Microsoft Graph where appropriate.
• Design instructions, prompts, topics, actions, knowledge grounding, variables, decision logic and exception handling for reliability and maintainability.
• Plan and execute functional testing, agent evaluation, user acceptance testing, error handling, security checks and performance monitoring.
• Iterate rapidly using evidence from tests and real users, while controlling scope and technical debt.
• Document architecture, requirements, data sources, permissions, dependencies, controls, ownership, support and lifecycle arrangements.

Adoption and value realisation

• Identify affected user groups and support practical change-impact assessment, communications and engagement.
• Create user guidance, deliver training and enable champions or citizen developers within agreed governance boundaries.
• Define adoption and outcome measures before launch, then monitor usage, completion, quality, rework, satisfaction and business value.
• Run post-pilot and post-implementation reviews; recommend whether to scale, improve or retire solutions.
• Share reusable patterns and lessons to improve the organisation’s AI and automation delivery capability.

Core deliverables

• Discovery plans, interview guides and workshop outputs.
• Current-state and future-state process maps.
• Evidence logs, problem statements, requirements, user stories and acceptance criteria.
• Use-case register, scoring model, prioritised roadmap and investment recommendations.
• Benefits hypothesis, baseline, business case and benefits-realisation plan.
• Solution options, MVP scope, design artefacts and governance assessment.
• Working agents, automations and integrations within the agreed low-code scope.
• Test plans, evaluation results, UAT evidence and production-readiness recommendations.
• Training, adoption, operational handover and continuous-improvement materials.

Required experience

• Substantial experience in business analysis, digital transformation, intelligent automation, AI consulting or a closely related role.
• Evidence of independently leading discovery with operational users, subject-matter experts and senior decision-makers.
• Strong process-modelling and requirements-engineering experience, including complex exceptions and cross-functional hand-offs.
• Experience assessing and prioritising technology or automation opportunities using transparent value, feasibility and risk criteria.
• Hands-on experience building business solutions with Microsoft Copilot Studio, Agent Builder and Power Automate.
• Experience taking solutions through discovery, design, pilot, testing, adoption and operational handover.
• Experience defining measurable outcomes and evaluating whether expected benefits were realised.
• Ability to work independently while recognising when architecture, security, data engineering or software-development specialists are required.

Desirable experience

• Client-facing consulting or managed-services experience.
• Microsoft 365 Copilot adoption, readiness, licence optimisation or use-case programmes.
• Power Apps, Dataverse, AI Builder, Azure AI services, Azure AI Search or Azure OpenAI.
• Microsoft Graph, custom connectors, REST APIs and common enterprise integration patterns.
• Application lifecycle management, environments, solutions, deployment pipelines and low-code centres of excellence.
• Agent evaluation, analytics, prompt optimisation, retrieval-augmented generation and knowledge-management design.
• Agile product delivery, backlog management and product-owner responsibilities.
• Benefits management, change management or service-transition experience.

Technical skills

• Core Microsoft stack: Microsoft 365 Copilot, Copilot Studio, Agent Builder, Power Automate, Power Platform, Teams and SharePoint.
• Data and integration: Microsoft Graph, connectors, APIs, Dataverse and an understanding of data quality, ownership and access controls.
• Azure AI: working awareness of Azure AI capabilities and when a low-code solution should escalate to a more engineered Azure pattern.
• Agent design: instructions, prompts, topics, actions, tools, grounding, orchestration, testing, monitoring and human hand-off.
• Delivery discipline: versioning, environments, application lifecycle management, test evidence, documentation and support readiness.
• Security and governance: identity, permissions, data protection, least privilege, auditability, responsible AI and operational controls.

Business and consulting skills

• Structured discovery, active listening and incisive questioning.
• Process mapping, root-cause analysis and future-state design.
• Business-case development, benefit estimation and option appraisal.
• Workshop facilitation and consensus building across technical and non-technical groups.
• Clear written communication for executives, delivery teams and end users.
• Commercial judgement, scope control and pragmatic prioritisation.
• Ability to explain AI capabilities, constraints, costs and risks without hype.

Behavioural competencies

• Outcome-led: starts with the problem and measurable result, not the technology.
• Constructively sceptical: tests assumptions, vendor claims and apparent use cases against evidence.
• Pragmatic: favours the smallest useful and supportable solution.
• Inclusive: involves affected users and designs for accessibility, fairness and different levels of digital confidence.
• Accountable: makes ownership, decisions, risks and unresolved assumptions visible.
• Curious: learns emerging capabilities while applying disciplined evaluation.
• Collaborative: connects business, technology, security, data and change perspectives.
• Resilient: is comfortable iterating when early hypotheses or prototypes do not succeed.